Understanding the Kinetics of Densified Activated Sludge: Implications in Design and Optimization

INTRODUCTION
For densification at full-scale continuous flow water resource recovery facilities (WRRFs), biological selection principles derived from selector design and aerobic granular sludge (AGS) concepts have been coupled with physical selection to lower the sludge volume index (SVI) and achieve superior sludge settling. In densified activated sludge (DAS), flocs co-exist with granules, and significant efforts have been made to research their settling behavior. However, the respective roles of flocs and different sized granules in biological nutrient removal kinetics and their impact on process performance and design of densified WRRFs remain unclear. The relative abundance and microbial activity distribution of ammonia oxidizing bacteria (AOB), nitrite oxidizing bacteria (NOB), and ordinary heterotrophic organisms (OHO) among flocs and granules and its influence on the process performance was investigated. This study focused on the kinetic contribution of different particle size-classes in DAS.

MATERIAL AND METHODS
DAS samples from the full-scale WRRF at Pueblo CO's James R. Dilorio WRRF (referred herein as Pueblo) and from the DAS demonstration treatment train at Metro Water Recovery's Robert W. Hite Treatment Facility (referred herein as Metro) were collected and characterized.

Size-class distribution of DAS was determined by sieving: flocs were 200 µm or smaller particles, small granules were 200-600 µm and large granules were 600 µm and larger. Particle size distribution analysis was conducted for each size-class.

Samples for microbial community analysis were collected from each size-class and stored at -18°C. Microbial community analyses were conducted with Illumina-based 16S rRNA amplicon sequencing. Quantitative PCR (qPCR) was performed on the DNA extracts to quantify total bacteria (16S). Species-level microbial communities were determined with long-read amplicon sequencing.

The biological kinetic rates for nitrification, denitrification and SND for each size-class were assessed through lab-scale studies.

RESULTS AND DISCUSSION
Nitrification Activity for Different Size-Class Distribution: Figure 1 presents the microscopy imaging for the three particle size classes investigated during this study for Metro. Specific nitrification rate (NH4RR) experiments were performed, since abundance and activity of nitrifiers among the different size-classes in DAS are not necessarily proportional. In general, the larger the size of the bio-aggregates, the slower their specific AOB and NOB activities (Figure 2). Small aggregates or flocs have higher specific NH4RR at DO concentrations ranging from 0.2 to 3.0 mg/L. Figure 3 depicts the maximum nitrification rates for the size-classes for Pueblo and Metro's DAS samples; the highest rates were observed for the small granules. In the case of Pueblo, the large granules offered the lowest nitrification rates. For Metro, the maximum rates for flocs and large granules were comparable. Therefore, the benefit of large granules is unclear when it comes to nitrification in DAS.

Figure 4 shows the observed half-saturation oxygen concentration for AOB (KDO, AOB) for the size-classes determined from lab tests. As the size of aggregate increases, there is greater diffusion resistance, so the higher their apparent KDO, AOB (similar results for NOB were observed). The KDO, AOB for Pueblo are generally lower than those estimated at Metro, as Pueblo operates at comparatively low DO concentrations.

Denitrification and SND Activity for Different Size-Class Distribution: The larger the size of the bio-aggregates, the smaller their specific denitrification rates (SDR). In the case of Metro, 4.58, 3.80, and 3.11 mg N/g MLVSS/hr were measured for flocs, small granules, and large granules, respectively. TIN removal efficiencies via SND for the bio-aggregate classes from Pueblo were measured and results are presented in Figure 5. Lower DO concentrations increased SND for all the size-classes. The highest SND efficiencies were observed at DO concentrations of 0.3 mg/L, 0.75 mg/L and 1.0 mg/L for flocs, small and large granules, respectively. For granules, SND reactions were nitrification-limited at low DO conditions, since the operating DO was much lower than the apparent KDO, AOB (reported in Figure 4). After maximum TIN removal efficiencies were achieved for each size-class, higher DO values decreased the TIN removal as the oxygen in the bulk liquid penetrated deeper inside the granules reducing the anoxic zone. A main challenge is finding a balance for aeration: to achieve full nitrification while also establishing anoxic zones in the core of the granules to preserve SND.

Microbial Ecology for Size-Class Distribution: Results will be included for the paper and presentation.

CONCLUSION
The study confirmed that diffusion and counter-diffusion resistance of substrates are inversely correlated with the size of the particles. Nitrification is maximized as particle size is decreased or DO is increased. SND is maximized in flocs when DO is decreased because nitrification is not limited, while it is maximized in small and large granules at higher DO. Controlling the size-class distribution in DAS must be considered for the design and operation of BNR systems, as optimized size will balance settleability and nutrient removal benefits.